Download HEAVY RAINFALL AND TEMPERATURE PROYECTIONS IN A

Artículo científico / Scientific paper
C IENCIAS DE LA T IERRA
pISSN:1390-3799; eISSN:1390-8596
UNIVERSIDAD
POLITÉCNICA
SALESIANA
http://doi.org/10.17163/lgr.n25.2017.02
H EAVY RAINFALL AND TEMPERATURE PROYECTIONS IN A
CLIMATE CHANGE SCENARIO OVER Q UITO , E CUADOR
P ROYECCIONES
DE LLUVIA Y TEMPERATURA EXTREMA EN ESCENARIOS DE
CAMBIO CLIMÁTICO SOBRE
Q UITO , E CUADOR
Sheila Serrano Vincenti1,∗ , Jean Carlos Ruiz2 and Fabián Bersosa3
1 Grupo
de Investigación en Ciencias Ambientales GRICAM, Centro de Investigación en Modelamiento Ambiental CIMA-UPS/
Universidad Politécnica Salesiana/Red de Universidades Frente al Cambio Climático y Gestión de Riesgos, Quito, Ecuador.
2 Escuela Politécnica Nacional/Red de Universidades Frente al Cambio Climático y Gestión de Riesgos, Quito, Ecuador
3 Grupo de Investigación en Ecología y Gestión de Áreas Protegidas, Centro de Investigación en Modelamiento Ambiental CIMAUPS/ Universidad Politécnica Salesiana/Red de Universidades Frente al Cambio Climático y Gestión de Riesgos, Quito, Ecuador.
*Autor para correspondencia: [email protected]
Manuscrito recibido el 14 de noviembre de 2016. Revisado el 14 de diciembre de 2016. Aceptado el 30 de diciembre
de 2016. Publicado el 31 de diciembre de 2016.
Abstract
This research analyzes daily extreme events of minimum, maximum temperatures and rain in the Metropolitan District of Quito using data of more than 30 years from the meteorological network of INAMHI (Instituto Nacional de
Meteorología e Hidrología de Ecuador) using the R- ClimDex computer program. A scenario for the year 2032 combining statistical results of extreme events and physical forcing from PRECIS scenarios A2 and B2 is also presented;
using the extreme value theory from extReme computer program. The results showed an increase in extreme minimum and maximum monthly temperature values in both, magnitude and frequency; and an increase in the intensity
of heavy rainfall. Projections to 2022 maintain this behavior, with results that should be taken into account by policy
makers and scientists due to the danger they mean for Quito’s ecosystem.
Keywords: extreme values, precipitation; temperature, Metropolitan District of Quito, climate change scenarios
16
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
Heavy rainfall and temperature proyections in a climate change scenario over Quito, Ecuador
Resumen
Esta investigación analiza eventos extremos a nivel diario de temperaturas mínimas, máximas y lluvias en el Distrito
Metropolitano de Quito utilizando datos con más de 30 años de la red meteorológica del INAMHI (Instituto Nacional de Meteorología e Hidrología de Ecuador), y utilizando el programa R-ClimDex. Se presentan escenarios el año
2032 combinando resultados estadísticos de eventos extremos con el forzamiento físico de los escenarios A2 y B2 del
modelo de cambio climático PRECIS A2 y B2, y utilizando la teoría de valores extremos del programa extReme. Los
resultados mostraron un incremento en los valores mensuales mínimos y máximos de temperatura tanto en magnitud
y frecuencia; además de un aumento en la intensidad de lluvias extremas. Las proyecciones para 2032 mantienen este
comportamiento, con resultados que deben ser tomados en cuenta por los tomadores de decisión y científicos debido
al peligro que significan para el ecosistema de Quito.
Palabras claves: valores extremos, precipitación; temperatura, Distrito Metropolitano de Quito, escenarios de cambio
climático.
Forma sugerida de citar:
Serrano, S., J. C. Ruiz and F. Bersosa. 2017. Heavy rainfall and temperature proyections in
a climate change scenario over Quito, Ecuador. La Granja: Revista de Ciencias de la Vida.
Vol. 25(1):16-32. pISSN:1390-3799; eISSN:1390-8596.
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
17
Artículo científico / Scientific paper
C IENCIAS DE LA T IERRA
1
Sheila Serrano Vincenti, Jean Carlos Ruiz and Fabián Bersosa
Introduction
The last report of the Intergovernmental Panel on
Climate Change (IPCC AR5, 2014) indicated that
there is a change in frequency and intensity of extreme weather events such as heat waves, intense precipitation, flooding, etc., in various regions of the
world as a result of global climate change. These
changes are particularly important for society and
the environment, since by definition, they are outside the range of usual ecosystem adaptability, and
thus can lead to severe impacts in biodiversity, agriculture, health infrastructure and economic losses
(García et al., 2012).
Particularly, a general increase of temperature
in Latin America was reported (Samaniego et al.,
2009). It was also registered an increase in temperature of one tenth of a degree per decade over the
Andes (Martínez et al., 2009); studies of Ecuadorian weather have shown that the temperature is
gradually increasing over the region (Vuille et al.,
2008; cited in Villacis et al., 2012). And studies have shown an increase of temperature in the four regions of Ecuador (Nieto et al., 2002; Cáceres, 1998).
On the DMQ Zambrano-Barragán et al. (2010), and
Villacis (2008), reveal an increase of annual temperature by 0.12◦C per decade over a period of the last
100 years.
1.1
Extreme events indexes
In addition to the gradual behavior of a variable
such as temperature, extreme events should also
be recorded. The CCI / CLIVAR / JCOMM (Expert Team on Climate Change Detection and Indices) proposed a methodology which includes the
RClimDex program for study extreme events in a
climate change scenario (Karl et al., 1999; Peterson,
2001).
Thus, using RClimDex over the Metropolitan
District of Quito DMQ, three possible threats related to climate change were identified: the extreme
values of maximum and minimum daily temperatures all over the region, and the intensity of rainfall
over the 90th percentile in the south and southwest
of the DMQ. This analysis was constructed using
available data from weather stations of INAMHI
(Instituto Nacional de Meteorología e Hidrología
de Ecuador) of the past 47-48 years (except station
Tomalón-Tabacundo) with 21 years, located in four
points of the District (Table 1). The trends are shown
in Table 2.
When studying extreme values, it is necessary
to change the statistical distribution because intense climate extreme events are more frequent
than expected by normal distributions (Gilleland
and Watts, 2005). The National Science Foundation
(NSF) through the National Center for Atmospheric Research (NCAR), the Weather and Climate Impact Assessment Science Initiative, and the NCAR
Geophysical Statistics Project (GSP), have recommended the use of the Extreme Value Theorem
(EVT) and developed tools for the study of extreme weather events included in a specific software
called eXtreme.
Table 1. Available weather stations with daily data selected for the study.† Cotopaxi-Clirsen station not presented continuous data.
Station
Izobamba
Papallacta
Tomalón-Tabacundo
† Cotopaxi-Clirsen
18
Code Station
Latitude
Longitude
Altitude (m.a.s.l.)
From
To
Temporal range
M003
0◦ 22’S
78◦ 33’W
3058
1964
2011
47
M188
0◦ 21’54”S
78◦ 8’41”W
3150
1963
2011
48
MA2T
0◦ 2’N
78◦ 14’W
2790
1990
2011
21
M120
0◦ 37’24”S
78◦ 34’53”W
3510
1964
2011
47
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
Heavy rainfall and temperature proyections in a climate change scenario over Quito, Ecuador
Table 2. Annual trends of each R-Climdex indicator of climate change for extreme temperatures and precipitation for the four
weather stations of DMQ and surroundings, ∗ mean values with a significance superior to 90 % (p<0.2).
INDEX
Izobamba
Papallacta
Tomalón-Tabacundo
Cotopaxi-Clirsen
M003
M121
M188
MA2T
0.03∗
0.095
0.067∗
0.051
0
0.404
0.033
0.243
0.01∗
–
0.031∗
0.125∗
0.11
0.001
Maximum daily minimum temperature [◦ C/year] (TNx)
p-value
Maximum daily maximun temperature (TXx) [◦ C/year]
p-value
0.2
Number of heavy precipitation
days (greater than 10 mm/day)
[day/year] (R10 mm)
0.16∗
0.049
-0.063
-0.037
p-value
0.135
0.756
0.796
0.863
Number of very heavy precipitation days (20 mm/day) (R20
mm) [day/year]
0.135∗
0.057∗
-0.127
-0.137∗
p-value
0.005
0.038
0.464
0.028
Table 3. Studied ecosystem classification in DMQ.
Station/Code
Ecosystem Classification
Altitudinal
Variation
(m.a.s.l.)
Minimum
Annual
Temperature (◦ C)
Maximum
Annual
Temperature (◦ C)
Annual
Precipitation
(mm)
Izobamba/M003
Artificial urban areas
2400-3100
10
16
960 (Murray, 1997)
Papallacta/M188
High montane evergreen
forest. Polylepis Upper
montane Andean north
forests Polylepis (Josse et
al., 2003)
4100-2900
6
17
922 (Baquero et
al., 2004; cited in
MECN, 2009)
TomalónTabacundo/
MA2T
Espinar
dry
montane
(Valencia et al., 1999),
Matorral semi humid montane forest (Valencia et al.,
1999) y (Baquero et al.,
2004)
2000-3000
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
5
18
575 (Josse et al.,
2003; cited in
MECN, 2009)
19
Artículo científico / Scientific paper
C IENCIAS DE LA T IERRA
1.2
Sheila Serrano Vincenti, Jean Carlos Ruiz and Fabián Bersosa
Ecosystem description of the sample re implemented by supercomputers, with horizontal resolutions of 300 km. In order to understand the
points
The DMQ is characterized by a great climatic and
orographic variety, with the northwestern tropical,
desert in the Guayllabamba Valley, the inter-andean
permanently clouded forest in the cold mountain
to urbanized city of Quito between the mountains
that surround it. Giving as result a wide variety of
ecosystems. However, due to the nature of the research, daily station data available only describe three types of ecosystems –with enough confidence–,
which are presented in Table 3.
1.3
Generalized Extreme Value Distribution
Let X1 , . . . , Xn be a sequence of independent random
variables, and let Mn = max{X1 , . . . , Xn } the maximum (or minimum) values measured on a regular
timeline, so Mn represents the extreme values of the
process in n time units of observation. For this data,
and using a linear renormalization, the distribution
of the set of Mn is given by the Generalized Extreme
Value (GEV), which has the form:
( 1 )
z−µ −ξ
(1)
G(z) = exp − 1 + ξ
σ
> 0 and (−∞ < µ < ∞), σ > 0,
where 1 + ξ z−µ
σ
(−∞ < ξ < ∞) are the parameters of location, scale
and shape respectively (Coles, 2004).
This distribution depends on the sign of ξ, if
ξ < 0 we have the Weibull distribution, usually associated with temperature data if ξ = 0, we have the
Gumbel distribution, and if ξ > 0 it is a Fréchet distribution, commonly used to simulate the behavior
of precipitation (García et al., 2012).
An advantage of GEVD is the possibility of a
non-stationary model, allowing a time-dependent
distribution by the parameter µ1 :
µ(t) = µ0 + µ1t.
(2)
The parameter µ1 corresponds to the change of
the location parameter, and could simulate the increasing or decreasing effects of climate change on
weather variables.
1.4
Climate change models
In order to simulate the behavior of climate change
in the planet, Global Circulation Models (GCM) we-
20
behavior of the weather at smaller scales, there are
Regional Climate Models (RCM), which work with
scales of 50 km or less, allowing for more precise
characteristics of the land surface and complicated
mountainous topography, coastlines and the inclusion of small islands and peninsulas.
RCMs are very complete dynamic models, based on the physics of the climate system, and virtually represent all processes, interactions and feedbacks between climate systems and the components
of the GCMs (PRECIS, 2004).
An example is PRECIS (Providing Regional Climates for Impacts Studies), which is a regional climate modeling system developed by the Hadley
Centre of the Met Office in the United Kingdom.
It is a free software that allows the use of highresolution data in impact, vulnerability and adaptation studies as recommended by the United Nations
Framework Convention on Climate Change (Articles 4.1, 4.8 and 12.1).
PRECIS works with HadCM GCM model,
which is forced by surface boundary conditions
such as sea surface temperature and sea-ice fraction. It has two time periods: 1960-1990 a base time or “control” period, used for comparisons with
real data; and 2000-2100, period used for forecasting. According to the IPCC (2001), circulation models are suggested to work with the A2 and B2 future scenarios agreeing to the economic and productive behavior of the planet and the possible incorporation of clean technologies. Where B2 scenario
describes a world with a greater emphasis on local solutions to economic, social and environmental
sustainability than A2. Both should be considered
equally right. The scenarios do not include additional climate initiatives. In this work, these two scenarios are used.
2 Materials and methods
2.1 DGVE, return levels and confidence intervals with real data
This study recorded maximum and minimum temperatures in four stations, which were the only available in to DMQ with sufficient temporal range (more than 30 years) and daily resolution; as shown
in Table 1, their behavior and trends were calculated using R-Climdex (Serrano et al., 2012), and then
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
Heavy rainfall and temperature proyections in a climate change scenario over Quito, Ecuador
used to calculate the climate change indexes shown
in Table 2. With these outputs, and using the extReme software and DGVE functions, future behaviors
were estimated of the maximum allowance for extreme events of rain and temperature, with return
levels for 2, 5, 10, 15 and 20 years and confidence
intervals of 95 %, for both A2 and B2 scenarios since
2012.
the intersection in the middle of the series and the
p-value for each case and achieving a correction factor by the difference in to the middle of the studied
series (Table 4).
3
Results
3.1
2.2 Fitting PRECIS data for each scenario
In order to determine the validity of PRECIS, its
gridded outputs for temperature, were used in the
“control” period from 1960 to 1990 and compared
with similar temporal periods for each station. A linear correlation for both the observed data and the
modeled by PRECIS was made, finding the slope,
Analysis of Maximum Temperatures by
meteorological station
3.1.1 Izobamba
In Figure 1 the behavior of the maximum monthly
value of the daily maximum temperature (TXX)
in Izobamba, analysis achieved by R-ClimDex is
shown. The trend of 0.01oC per year is statistically
significant at 76.3 % (Serrano et al., 2012).
Table 4. Correction factor by available meteorological station, between PRECIS time series and observed data, for temperature.
Observed data
Control PRECIS
Correction factor
Station
Slope
p-value
Intersection
Slope
p-value
Intersection
PRECIS-Observed data
M003
1,08E-05
0,0089
11,93
4,13E-05
8,27E-67
11,64
0,294970646
M180
3,92E-05
4,27E-14
9,76
4,34E-05
2,47E-45
10,11
-0,352601947
M120
1,6693E-04
2,19E-22
11,93
4,50E-05
1,45E-72
10,16
-0,712912888
Figure 1. Behavior of annual maximum of daily maximum temperature (TXX) in Izobamba.
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
21
Artículo científico / Scientific paper
C IENCIAS DE LA T IERRA
Sheila Serrano Vincenti, Jean Carlos Ruiz and Fabián Bersosa
Figure 2. Adjusting of the maximum annual value of daily maximum temperature (TXX) in Izobamba for DGVE Weibull type
distribution. The first two upper graphs show the proper fit of the model, while the lower left and right graphs show the return
periods with confidence limits of 95 % (blue line) and the probability density distribution.
Figure 2 shows the data set to a covariant DGVE where the results, after the maximum likelihood
method, indicate that the location parameter varies
over time like: µ = 22,65353(0,11592) + 0,01t the scale parameter is σ = 0,65709(0,08777), and the shape parameter is ξ = −0,36045(0,14134), following a
Weibull distribution. The maximum likelihood was
0.01303611, and the p-value has a significance value
above 98 %. Values in parentheses indicate the standard deviations of each parameter.
Return values and confidence limits for the
22
Izobamba data studied up to 100 years are presented in the lower part of Figure 2, but in detail, and for 2022 year are shown in Table 5.
The value of the shape parameter for all periods
is ξ = −0,3604(−0,61849, −0,10241). Also, the settings of trend data achieved with PRECIS for A2
and B2 scenarios are presented in Table 4 The
value of the shape parameter for all periods is
ξ = −0,1751(−0,31983, 0,02098) to A2; and ξ =
−0,2264(−0,40246, −0,03666) to B2.
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
Heavy rainfall and temperature proyections in a climate change scenario over Quito, Ecuador
Table 5. Return periods, return levels and confidence intervals at 95 % for the actual data of maximum temperatures in Izobamba.
Return
Observed
PRECIS A2
PRECIS B2
period
Return level
LI
LS
Return level
LI
LS
Return level
LI
LS
(years)
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
2014
22.879
22.652
23.1144
23.133
22.91
23.3696
23.168
22.8978
23.4483
2017
23.414
23.1986
23.6392
23.807
23.551
24.1132
23.9056
23.6147
24.2505
2019
23.666
23.4628
23.9786
23.999
23.7303
24.3542
24.1082
23.8077
24.5028
2022
23.781
23.5809
24.1296
24.185
23.9005
24.6095
24.2996
23.9878
24.7644
2027
23.851
23.6515
24.2306
24.377
24.0736
24.9012
24.494
24.1674
25.0563
2032
22.879
22.652
23.1144
24.503
24.1851
25.0895
24.6195
24.2809
25.2437
Table 6. Return periods, return levels and confidence intervals at 95 % for analyzed data and both A2 and B2 scenarios of PRECIS,
for the maximum annual temperatures in Tomalón-Tabacundo.
Return
Observed
PRECIS A2
PRECIS B2
period
Return level
LI
LS
Return level
LI
LS
Return level
LI
LS
(years)
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
2014
27.7286
27.1488
28.3307
28.5509
28.2998
29.5172
28.0392
27.4989
28.0392
2017
28.6298
28.0676
29.3502
29.2193
29.0496
29.5375
28.7569
28.2599
29.3245
2019
28.858
28.3054
29.7135
29.3278
29.2666
29.6931
28.9228
28.4614
29.5541
2022
29.0654
28.5221
29.9473
29.4081
29.2859
29.825
29.0671
28.6424
29.67
2027
29.2674
28.7319
30.2188
29.4709
29.4015
29.7692
29.2013
28.8123
29.8465
2032
29.3929
28.8606
30.4183
29.5028
29.4487
29.7336
29.2812
28.9124
29.9996
3.1.2 Tomalón-Tabacundo
3.1.3 Papallacta
The same statistical treatment as performed to Izobamba was applied, and was recorded the highest
slope of DMQ (0.125◦C / year) with a statistical significance of 0.999 %. By simulating the data with a
covariant DGVE the maximum likelihood method
indicate a temporal variation of the location parameter: µ = 27,35713(0,29314) + 0,125t, the scale parameter was σ = 1,07630(0,21311), and the shape
parameter was ξ = −0,33074(0,19109), showing that
these data follow a Weibull distribution; the p-value
was 0.1150664. The return levels and confidence limits of the observed data and for the A2, B2 scenarios is presented in Table 6.
In the serie of the maximun value of maximun temperature there is a positive trend of 0.031◦C/year,
with a statistical significance of 0.89 %. The simulation a covariant DGVE and using the maximum
likelihood method indicate that the location parameter varies over time as µ = 19,39455(0,26281) +
0,031t, the scale parameter σ = 0,96090(0,19836), the
shape parameter ξ = −0,28744(0,22732), showing
that these data follow a Weibull distribution, the
value-p-value 0.2363144, significant at 76 % level.
Comparison of return levels and confidence intervals for the two scenarios and the observed data is
presented in Table 7.
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
23
Artículo científico / Scientific paper
C IENCIAS DE LA T IERRA
Sheila Serrano Vincenti, Jean Carlos Ruiz and Fabián Bersosa
Table 7. Return periods, return levels and confidence intervals at 95 % for analyzed data and both A2 and B2 scenarios of PRECIS
for the next 20 years, for the maximum annual temperatures in Papallacta.
Return
Observed
PRECIS A2
PRECIS B2
period
Return level
LI
LS
Return level
LI
LS
Return level
LI
LS
(years)
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
2014
19.7288
19.2086
20.2856
20.0045
19.4826
20.6801
20.4026
19.9515
20.928
2017
20.5654
20.0355
21.2828
21.231
20.4309
22.7266
21.2677
20.7114
22.1982
2019
20.7845
20.2607
21.5991
21.6826
20.75
23.5557
21.5366
20.9423
22.5901
2022
20.9868
20.468
21.8494
22.1763
21.0762
24.5801
21.8067
21.1655
23.0488
2027
21.1874
20.6696
22.1613
22.7625
21.4322
25.9559
22.0995
21.3945
23.6352
2032
21.314
20.7932
22.4003
23.1975
21.6749
27.083
22.2998
21.5419
24.0935
Figure 3. Maximum annual daily temperatures expected for the next 10 years in the DMQ in the color bar. Horizontal data shows
the Longitude, Vertical the Latitude. a) forecast using the observed trend with real data from the studied meteorological stations,
b) forecast using the product of dynamic forcing trend calculated by PRECIS A2 scenario c) Forecast using the product calculated
by the dynamic forcing B2 (optimistic) scenario of PRECIS trend.
24
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
Heavy rainfall and temperature proyections in a climate change scenario over Quito, Ecuador
The GCM simulated data were regionalized
using the technique of approaching averages, taking into account the annual mean temperature
maps of the Ministry of Environment (MDMQ,
2011) and PRECIS model outputs, thus maps are
presented with the resolution of this latter model. In
Figure 3(a) shows the highest values of annual maximum temperatures that can be expected to 2032
year, these results take into account the dynamic
forcing calculated by PRECIS scenarios A2 and B2,
which are shown in Figure 3(b) and 3(c) respectively.
3.2 Analysis of daily minimum temperatures for Izobamba, Tomalón-Tabacundo
y Papallacta
The behavior of the annual maximum values of minimum temperatures (early hours of dawn) for Izobamba station, located south of the DMQ is presented, into the series there is a positive trend of
0.03◦C/year with a statistical significance of 75.7 %.
By simulating the data with a covariant DGVE,
the location parameter was µ = 10,05092(0,13680) +
0,03t, the scale parameter was σ = 0,78644(0,09780),
and the shape parameter ξ = −0,22174(0,11901).
Showing that these data follow a Weibull distribution. Also the test of maximum likelihood indicates
that the p-value was 0,09103756.
In Tomalón-Tabacundo the trend was
0.051◦C/year, and the DGVE has a location parameter of µ = 14,38173(0,29938)+ 0,051t, a scale parameter of σ = 1,07082(0,23624), and a shape parameter
of ξ = −0,27288(0,27655), in a Weibull distribution,
the p-value was 0.3372716. Since in Papallacta, the
trend is positive too: 0.067◦C/year with a significance of 96.7 %. The return levels and confidence
intervals for the two scenarios and the observed data for the three stations for 2032 year is presented in
Table 8.
Similarly, the regional data of annual maximum
values of minimum temperatures in the DMQ, for
real data and the A2, B2 scenarios are presented in
Figure 4.
20 mm/day) for Izobamba station, located south of
the DMQ, shows a positive trend of 0.366 mm/year
with a statistical significance of 96.7 % (Figure 5).
However, in the case of precipitation, there was
not detected a direct dynamic forcing producing its
increase over time (IPCC, 2014), but the increase
in extreme weather events in general. That is why
we did not use PRECIS scenarios. Return levels and
confidence intervals are calculated only with the observed data, as shown in Table 9.
By simulating covariant DGVE results achieved after the maximum likelihood method in Izombamba, indicates that the location parameter varies
µ = 52,16587(2,98138) + 0,366t, the scale parameter σ = 17,45288(2,34374), the shape parameter ξ =
0,18686(0,11512), showing that these data follow a
Frechtel distribution with a p-value of 0.0622325.
The results are shown in Table 9(a).
In Tomalón-Tabacundo, the dryest region,
the maximun precipitation has a negative trend
of -0.161 mm/year with a p-value of 0.649,
the location parameter was chosen as µ =
27,44210(2,31995), with a σ = 8,65550(1,98911), and
ξ = 0,33574(0,22768), into a Frechtel distribution,
the p-value was 0.02581635 (Table 9(b)).
In Papallacta there is another negative trend of
-0.473 mm/yer (p-value of 0.211). The DGVE has
a location parameter of µ = 52,81604(7,64252), σ =
34,79840(6,15202), and ξ = 0,24224(0,14866), in a
Frechtel distribution with a p-value of 0.03283383
(Table 9(c)).
The regional data in Table 7 for maximum value
per year of heavy precipitation into DMQ is presented in Figure 6.
4
4.1
Conclusions and discussion
Behavior of temperatures and ecosystemic impacts in the DMQ
3.3 Behavior of the heavy rainfall in Izo- In the DMQ were identified three types of threats
bamba, Tomalón-Tabacundo and Pa- related to climate change or climate variability: a
statistically significant increase in the magnitude
pallacta
of both maximum and minimum temperatures and
The behavior of maximum annual values of maxi- an increase in the frequency of heavy rainy days
mum precipitation days (above the 95th percentile = (Serrano et al., 2012).
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
25
Artículo científico / Scientific paper
C IENCIAS DE LA T IERRA
Sheila Serrano Vincenti, Jean Carlos Ruiz and Fabián Bersosa
Figure 4. Annual maximum of daily Minimum Temperatures expected for 2032 year in the color bar. Horizontal data shows the
Longitude, Vertical the Latitude. a) forecast using the observed trend of real data from the meteorological stations, b) forecast
using the product of dynamic forcing trend calculated by the A2 scenario PRECIS, c) forecast using the trend product calculated
by the dynamic forcing B2 scenario of PRECIS.
Figure 5. Annual behavior of the daily maximum precipitation values recorded since 1964 in the year 2011 in Izobamba.
26
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
Heavy rainfall and temperature proyections in a climate change scenario over Quito, Ecuador
a) Izobamba
Return
Observed
PRECIS A2
PRECIS B2
period
Return level
LI
LS
Return level
LI
LS
Return level
LI
LS
(years)
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
2014
10.327
10.053
10.6168
10.5158
10.2428
10.7971
10.7593
10.465
11.0544
2017
11.054
10.7566
11.4034
11.2254
10.9413
11.546
11.4875
11.2043
11.788
2019
11.254
10.9494
11.6696
11.4139
11.1264
11.7798
11.6688
11.3901
11.9941
2022
11.444
11.129
11.958
11.5891
11.2972
12.0264
11.8321
11.5578
12.2047
2027
11.637
11.3068
12.2385
11.7641
11.465
12.2838
11.989
11.7192
12.4346
2032
11.761
11.418
12.4268
11.8753
11.5697
12.4345
12.087
11.818
12.5507
b) Tomalón-Tabacundo
Return
period
Observed
Return level
◦
PRECIS A2
LI
◦
LS
◦
Return level
◦
PRECIS B2
LI
◦
LS
◦
Return level
◦
LI
◦
LS
◦
(years)
[ C/day]
[ C/day]
[ C/day]
[ C/day]
[ C/day]
[ C/day]
[ C/day]
[ C/day]
[ C/day]
2014
14.7552
14.16747
15.40656
14.7282
14.04727
15.50257
14.9877
14.30136
15.67958
2017
15.6998
15.09104
16.54288
16.0592
15.22647
17.44137
16.055
15.39729
16.82908
2019
15.95
15.35243
16.86408
16.4696
15.58255
18.02369
16.3168
15.68436
17.25251
2022
16.1824
15.59416
17.16913
16.8801
15.92611
18.6907
16.5512
15.94356
17.48822
2027
16.4141
15.82815
17.58124
17.3233
16.27861
19.5283
16.7759
16.19098
17.76637
2032
16.5611
15.96987
17.90907
17.6253
16.50622
20.17527
16.9135
16.3407
17.9754
c) Papallacta
Return
Observed
PRECIS A2
PRECIS B2
period
Return level
LI
LS
Return level
LI
LS
Return level
LI
LS
(years)
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
[◦ C/day]
2014
9.098
8.51747
9.09802
11.3319
9.92041
11.33188
9.8874
9.17342
10.67584
2017
9.8408
9.29491
10.59829
12.9631
11.83807
14.12185
10.9109
10.12034
12.13227
2019
10.0219
9.4964
10.79966
13.312
12.27014
14.68806
11.1987
10.40246
12.48641
2022
10.1835
9.67933
10.96891
13.6047
12.64676
14.92382
11.4739
10.6679
12.9048
2027
10.338
9.855
11.19093
13.8665
12.99694
15.09987
11.7569
10.93114
13.45406
2032
10.4323
9.96152
11.36773
14.0171
13.20499
15.22524
11.9416
11.09534
13.88822
Table 8. Return periods, return levels and confidence intervals at 95 % for both scenarios A2 and B2 of PRECIS for the next 20
years, for the minimum annual temperatures in a) Izobamba b) Tomalón-Tabacundo c) Papallacta.
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
27
Artículo científico / Scientific paper
C IENCIAS DE LA T IERRA
Sheila Serrano Vincenti, Jean Carlos Ruiz and Fabián Bersosa
a)
Return period [years]
Year
Return level [mm/day]
IL 95 % [mm/day]
SL 95 % [mm/day]
2
2014
58.7867
52.3806
66.39552
5
2017
82.381
71.81226
98.9752
7
2019
91.2272
78.53774
113.82874
10
2022
100.9893
85.58649
131.11258
15
2027
112.6984
93.54083
152.2947
20
2032
121.4659
99.1595
169.26421
Return period [years]
Year
Return level [mm/day]
IL 95 % [mm/day]
SL 95 % [mm/day]
2
2014
30.8179
25.83315
37.5405
5
2017
44.3192
35.40184
61.89684
7
2019
49.9595
38.958
73.53879
10
2022
56.5417
42.74972
88.82883
15
2027
64.9244
19.16253
110.68628
20
2032
71.5448
13.47774
71.54475
Return period [years]
Year
Return level [mm/day]
IL 95 % [mm/day]
SL 95 % [mm/day]
2
2014
66.1534
49.78686
86.86547
5
2017
115.7551
87.95392
167.84057
7
2019
135.1206"
101.41638
201.95179
10
2022
156.9403
115.64823
243.90614
15
2027
183.7016
131.87326
300.17372
20
2032
204.1415
143.44968
143.44968
b)
c)
Table 9. Return periods, return levels and confidence intervals at 95 % for the actual data of maximum heavy rainfall in a)
Izobamba, b) Tomalón-Tabacundo, c) Papallacta.
28
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
Heavy rainfall and temperature proyections in a climate change scenario over Quito, Ecuador
Figure 6. Possible values of maximum daily precipitation forecasts for the next 10 years in the DMQ year in the color bar.
Horizontal data shows the Longitude, Vertical the Latitude. The figure shows the values of extreme events by day expected during
this period, since Table 7.
The first part of this study, was aimed to know
the intensity of extreme events of the minimum and
maximum temperatures in the near future (2032
year), using DGVE distributions which indicated
that the best fit to the temperature values was given
by a Weibull covariant distribution i.e. involving a
forcing behavior data. Thus, we have worked with
the atmospheric forcing presented in PRECIS for A1
and B1 scenarios. In consequence, this study used
both: a dynamic and statistical prediction.
Results show that for 2032 year, in the southern area (Izobamba), where the average daily maximum temperature is 14.6◦C, it will be possible observe extreme events such as temperatures up to
23.7◦C, and according to the A2 and B2 scenarios
temperatures as high as 24.3◦C and 24.2◦C respectively could be observed, i.e. an eventual increase
of about 10◦ over the average. It is expected to occur in the months of August and September, which
corresponding to the dry season in Quito.
In the southeastern region (Papallacta) where
the average daily maximum temperature is 14.4◦C,
it is expected to register extreme events of 21◦ C, and
according to the A2 and B2 models 21.8◦C and 21.1
respectively, about 7◦ C above average temperature. While in the north-east, the warmest region in
DMQ, averaging 21.6◦C, in 10 years it is possible to
find extreme values up to 29◦C and using the A2
and B2 scenarios of 29.4◦C, and 29◦ C respectively,
i.e. 8◦ C more than average.
In the case of minimum temperatures, in Izobamba whose average daily minimum temperatures are 6◦ C, is expected to record extreme events
of 11.4◦C, 11.6◦C and 11.8◦C according to statistical data and scenarios A2 and B2 respectively, an
increase of nearly 7◦ . In Papallacta, with an average lowest temperature of 5.5◦ C, it is expected to have extreme temperatures as low as 10.1◦C and A2
and B2 respectively 13.6◦C and 11.4◦ C models, an
increase of more than 8◦ C. Also, in the warmest re-
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
29
Artículo científico / Scientific paper
C IENCIAS DE LA T IERRA
Sheila Serrano Vincenti, Jean Carlos Ruiz and Fabián Bersosa
gion of the DMQ, where station Tabacundo Tomalon is located, has an average of 9.11◦C, and the model can be registered 16.2◦C, 16.8◦C and 16.5◦C according to statistical trend and the two scenarios A2
and B2, i.e. more than 7◦ C for the next 10 years.
It should be noted that these values are possible extreme temperature events, whose frequency
is casual, but the intensity has been shown to consistently rise may experience extreme events between 6 and 10◦ C more than average temperatures
to which ecosystems are already accustomed.
4.2
Behavior of heavy precipitation
In the case of the precipitation DGVE indicate that
the better option is to use a no covariant Fretchel
distribution. In Izobamba, registered as the rainiest
region, with a mean daily rainfall of 6.8 mm, it is expected to register single events to up 100 mm/day,
14 times more than the average. The southeastern
area represented by Papallacta has daily rainfall
averages 3.7 mm/day, and in the next 10 years may
register 156 mm/day, i.e. 42 times more. This phenomena occurs because Papallacta has historically
registered record rainfall of 183 mm/day and 160
mm/day on 27 and 25 March 2003 respectively. Therefore it is very likely that the coming years will see
a similar events of this magnitude.
As for the northeast, where the station TomalonTabacundo is located, average rainfall of 4.3
mm/day and can record maximum of 56.5 mm/day
are recorded. It should be noted that in this area the
rains are often scarce, and there is a tendency to decrease precipitation and hence the hardening of dry
regime.
4.3
Ecosystem impact
Because the DMQ have varied ecosystems with well
defined characteristics, it becomes clear that these
areas are affected differently.
On what it refers to the increase in minimum
temperatures, even though it is a measure generally
taken in the early hours of the morning, it is a direct
measuring of warm nights. And it would be an indicator of potentially harmful effects by the lack of
night cooling and main contributor of heat stress in
animals and plants, especially those located in the
transition zone of the paramo and are not adapted
to these conditions. Also, it is noted that besides the
presence of greenhouse gases that increase nightti-
30
me temperatures, must be added the heat island effect, which must be taken into account in all sampling points, as the effects urbanization resulting in
increased heat released during the night by modern
infrastructure.
The increase in maximum temperatures, usually
achieved at noon, directly affecting the adaptability
of animal and plant species, including humans; because in a context of physical stress, coupled with
high temperatures can be triggered death. Also, this
indicator can also be interpreted as a measure of
greater or lesser heliophany cloud cover, which can
also favor the drought (Frich, 1999).
Also in the context of climate change, greenhouse gases favor the hydrological cycle and collaborating nucleation of water vapor into rain. While the
increase of temperature favors the greatest amount
of water vapor available, and hence is generated
more intense precipitation. Those ecosystems that
have more absorbency due to its vegetation cover
would not be as affected by these sporadic events,
although the edges urbanized city located in areas
at risk by landslides are. Furthermore, it is noted
that in the dry Northeast of DMQ has identified a
negative trend in the presence of rain, which could
result in the tightening of dry conditions in the
area, affecting mainly to their biodiversity (Riebeek,
2005).
On the behavior of disease vectors, investigations of Rodriguez and Buitrón (20014) had been established that the increase of temperature and humidity favors the occurrence of diseases produced
by insect vectors such as Anopheles mosquitoes and
Aedes, responsible for the transmission of malaria
and dengue. It was reported in 2010, in the provinces of Carchi (300- 4.723 m.a.s.l.) and Imbabura
(1200 a 3000 m.a.s.l.) seven cases of dengue fever
were confirmed, since National Institute of Hygiene Izquieta Perez (INHIP). Aedes currently aegypti
is able to survive between 1,500 and 1,700 meter,
meanwhile Varsovia Cevallos (El Comercio, 2010)
conclude that under certain conditions of microclimate the mosquito could adapt to Quito and therefore could be cases of dengue, as happened in Galapagos in 2002 where the disease first and then the
vector was reported. Ramirez et al. (2009), estimated
that 34 % of the world population would be at risk
of contracting dengue. In Ecuador 70 % of its territory is favorable for the presence of Aedes aegypti.
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
Heavy rainfall and temperature proyections in a climate change scenario over Quito, Ecuador
Acknowledgements
This research was funded by the Climate Development Knowledge Network-CDKN within the Vulnerability Study DMQ. And it was done under the
direction of Stokolm Eviroment-SEI Institute, the
University Network for Climate Change represented by the Polytechnic National-EPN, the Pontifical
Catholic University the Ecuador-PUCE and the Research Modeling Environment Center CIMA-UPS
Salesian Polytechnic University. It also appreciated
the product management and validation of the Secretariat of the Environment of Illustrious Metropolitan District of Quito, and to MAE (Ministry of Environment) an INAMHI for the data.
5 References
Baquero, F., R. Sierra, L. Ordoñez, M. Tipán, L. Espinoza, M. Ribera and P. Soria. 2004. La Vegetación de los Andes del Ecuador. Memoria
explicativa de los mapas de vegetación potencial y remanente de los Andes del Ecuador a
escala 1:250.000 y del modelamiento predictivo con especies indicadoras.
Cáceres, L., G. Mejía and Ontaneda. 1998. Evidencias del cambio climático en Ecuador. Bull.
Inst. fr. Étudesandines. 27(3):547-556.
Coles, Stuart. 2004. S-plus functions for extreme value modeling: An accompaniment to the book an introduction to
statistical modeling of extreme values.
http://www.stats.bris.ac.uk/masgc/ismev/u
ses.ps
Gilleland, Eric and Richard W. Katz. 2005. Extremes Toolkit (extRemes): Weather and Climate Applications of Extreme Value. National Science Foundation (NSF) through the
National Center for Atmospheric Research
(NCAR) Weather and Climate Impact Assessment Science Initiative, with additional support from the NCAR Geophysical Statistics
Project (GSP).
IPCC. 2001. Climate Change 2001: The Scientific
Basis. Contribution of Working Group I to the
Third Assessment Report of the Intergovernmental Panel on Climate Change [Houghton,
J.T., Y. Ding, D.J. Griggs, M. Noguer, P.J. van
der Linden, X. Dai, K.]
IPCC. 2014. Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and
III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Core
Writing Team, R.K. Pachauri and L.A. Meyer
(eds.)]. IPCC, Geneva, Switzerland. page 151.
Josse, C., G. Navarro, P. Comer, R. Evans, D. FaberLangendoen, M. Fellows, G. Kittel, S. Menard,
M. Pyne, M. Reid, K. Schulz, K. Snow and
J. Teague. 2003. Ecological Systems of Latin
America and the Caribbean: A Working Classification of Terrestrial Systems. NatureServe.
Arlington, VA.
Karl, T., N. Nicholls and A. Ghazi. 1999. Clivar/gcos/wmo workshop on indices and indicators for climate extremes: Workshop summary. Climatic Change. 42:3-7.
El Comercio. 2010. El mosquito del dengue sí se adapta a Quito. 22 de marzo.
http://www.elcomercio.com/actualidad/mo
squito-del-dengue-adapta-quito.html
Martínez, R., D. Ruiz, V. Marcos, M. Andrade, L.
Blacutt, P. Daniel, et al. 2009. Synthesis of the
climate of the tropical Andes.
Frich, P. 1999. REWARD–A Nordic Collaborative Project. Annex of Meeting of the Joint
CCl/CLIVAR Task Group on Climate Indices,
Bracknell, UK, 2-4 September 1998. World Climate Data and Monitoring Programme.
MDMQ-Secretaría de Ambiente. 2011. Memoria
Técnica del Mapa de Cobertura Vegetal del
Distrito Metropolitano de Quito (DMQ). Quito.
García, O. Rafael Cueto, N. Santillan Soto, S. Ojeda
Benitez and M. Quintero Nuñez. 2012. Escenarios de temperaturas extremas en Mexicali,
México bajo condiciones de cambio climático.
páginas 349-358.
MECN. 2009. Ecosistemas del Distrito Metropolitano de Quito (DMQ). Serie de Publicaciones
del Museo Ecuatoriano de Ciencias Naturales
(MECN)–Fondo Ambiental del MDMQ. Publicación Miscelánea (6):1-51. Imprenta Nuevo
Arte. Quito, Ecuador.
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.
31
Artículo científico / Scientific paper
C IENCIAS DE LA T IERRA
Sheila Serrano Vincenti, Jean Carlos Ruiz and Fabián Bersosa
Murray, Sharon. 1997. Urban and Peri-Urban Forestry in Quito, Ecuador: a Case-Study. Forestry Department. Food and Agriculture Organization of the United Nations. Rome.
Nieto, J., R. Martínez, J. Regalado and F. Hernández. 2002. Análisis de tendencias de series de
tiempo oceanográficas y meteorológicas para
determinar evidencias de cambio climático en
la costa del Ecuador. Acta oceanográfica del
Pacífico. 11(1):17-21.
Peterson, T. 2001. Report on the activities of the
working group on climate change detection
and related rapporteurs 1998-2001. WMO,
Rep. WCDMP-47, WMO-TD 1071. página 143.
Geneve, Switzerland.
PRECIS. 2004. The PRECIS Handbook, Generating High Resolution Climate Change Scenarios using PRECIS. This handbook was published jointly by the UNDP and the Hadley Centre.
Ramírez, M. G. Zepeda, H. E. VelascoMondragón, C. Ramos, J. E. Peñuelas, et al.
2009. Caracterización clínica y epidemiológica
de los casos de dengue: experiencia del Hospital General de Culiacán, Sinaloa, México.
Rev. Panamá Salud Pública 2009. 25:16-23.
Riebeek, Holli. 2005. The Rising Cost of Natural
Hazards. Earth Observatory.
Rodríguez, A. and M. Buitrón. 2015. Enfermedades sensibles al clima en el Distrito Metropolitano de Quito, un análisis temporal en
el periodo 2001-2010. La Granja: Revista de
Ciencias de la Vida. 21(1):16-33.
Samaniego, J. 2009. Cambio climático y desarrollo en América Latina y el Caribe: una reseña.
CEPAL, Santiago, documento de Proyecto.
32
Serrano, S., D. Zuleta, V. Moscoso, P. Jácome, E.
Palacios y M. Villacís. 2012. Análisis estadístico de datos meteorológicos mensuales y diarios para la determinación de variabilidad climática y cambio climático en el Distrito Metropolitano de Quito.La Granja. 16(2):23-47.
ISSN: 1390-3799.
United Nations. 1995. Framework Convention on
Climate Change. Conference of the parties.
First session. Berlin.
Valencia, R., C. Cerón, W. Palacios and R. Sierra. 1999. Las Formaciones Naturales de la
Sierra del Ecuador. Propuesta Preliminar de
un Sistema de Clasificación de Vegetación para el Ecuador Continental. Proyecto INEFAN/
GEF-BIRF y EcoCiencia. páginas 79-108. Quito, Ecuador.
Villacís, M., F. Vimeux and J. Taupin. 2008. Analysis of the climate controls on the isotopic composition of precipitation (d18o) at nuevo rocafuerte, 74.5◦W, 0.9◦S, 250 m, Ecuador. Comptes
Rendus Geoscience. 340:1-9.
Villacís, M., A. Fernández, J.C. Pouget and M. Escobar. 2012. Impactos del cambio climático en
el sector agua durante los últimos 30 años se
identificación de los aspectos que constituyen
la vulnerabilidad. Estudio de Vulnerabilidad
del DMQ (not published).
Zambrano-Barragán, C., O. Zevallos, M. Villacís
and D. Enriquez. 2010. Quito’s climate change
strategy: A response o climatic change at the
metropolitan district of Quito, Ecuador. Resilient Cities: Cities and Adaptation to Climate
Change.
L A G RANJA :Revista de Ciencias de la Vida 25(1) 2017:16-32.
c
2017,
Universidad Politécnica Salesiana, Ecuador.